基于SV模型的隔夜信息对股票收益和波动影响的研究

发布时间：2018-06-01 01:19

本文选题：隔夜信息 + 资产收益波动模型　；参考：《东北大学》2014年硕士论文

【摘要】：由于经济一体化的发展和金融全球化的大趋势,在世界各个国家,金融市场间的关联越来越密切,因此对全球宏观经济信息所作出的反应也逐渐趋同。任何一个可能会使得证券资产价格产生波动的信息在这种全球化体系的背景下,都能够在全球范围内的资本市场上有效而快速的扩散。虽然这种信息在不同地区的市场上体现出不同程度的影响,但是任何一个市场都不应该忽略这种外界信息的存在。由于时差原因,我国大陆股市与主要的欧美股票市场无法同步交易,而欧美股市的主要交易信息,都在我国闭市期间产生了累积。作为一个新兴的市场,如何削减中国A股市场激烈的价格波动,一直为相关监管部门、学术机构研究者和投资人员所关心。中国大陆的A股市场作为全球化经济大家庭当中的一部分,国际主要股票市场对它产生的影响越来越深。本文以沪深指数为研究数据,研究了夜间信息对股市收益与波动的影响。本文比较了两种典型的金融波动模型,选取了刻画能力较好的基于T分布的SV模型,并且在基础SV模型的均值和波动等式中加入了滞后一期的收益项、隔夜信息项并对隔夜信息进行分类,在波动等式中加入不对称波动。并选取沪深指数为研究数据,采用MCMC方法对模型进行估计,根据估计结果验证本文提出的三个假设。本文的研究结论主要包括：第一,基于T分布的SV模型对金融时间序列分布的“高峰厚尾”和平方序列的长记忆性有更好的刻画能力；第二,日内和隔夜收益是两种不同的数据来源,并且隔夜信息对白天的收益和波动有一定的预测能力；第三,在不考虑隔夜期间的长度的情况下,不同类型的隔夜信息有不同的预测能力,对隔夜信息进行分类是有必要的；第四,股票市场中存在着杠杆效应,且日内信息和隔夜信息产生的杠杆效应不同。对于白天的信息,负的冲击会引起更大的波动,且正负冲击都能够增加波动。而对于夜间信息来说,市场不同,隔夜信息种类不同,捕捉到隔夜信息对日内波动的影响也就不相同。
[Abstract]:With the development of economic integration and the trend of financial globalization, the relationship between financial markets is becoming more and more close in every country of the world, so the response to global macroeconomic information is gradually converging. Under the background of this kind of globalization system, any information that may make the securities asset price fluctuate can spread effectively and rapidly in the global capital market. Although this kind of information reflects different degrees of influence in different regional markets, no market should ignore the existence of this kind of external information. Because of the time difference, the mainland stock market in China and the major European and American stock markets can not trade synchronously, and the main trading information of the European and American stock markets has accumulated during the closing period of our country. As an emerging market, how to reduce the fierce price volatility in China's A-share market has been concerned by relevant regulators, academic researchers and investors. As part of the global economic family, the A-share market in mainland China has a growing influence on the major international stock markets. In this paper, the influence of night information on stock market returns and volatility is studied based on the data of Shanghai and Shenzhen Index. In this paper, two typical financial volatility models are compared, and the SV model based on T distribution is selected, and the return term with lag period is added to the mean value and volatility equation of the basic SV model. The overnight information items are classified and asymmetric fluctuations are added to the fluctuation equation. The Shanghai and Shenzhen indices are selected as the research data and the MCMC method is used to estimate the model. According to the estimation results, the three hypotheses proposed in this paper are verified. The main conclusions of this paper are as follows: first, the SV model based on T distribution has better characterizing the "peak thick tail" of financial time series distribution and the long memory property of square sequence; second, Intra-day and overnight earnings are two different sources of data, and overnight information has a certain ability to predict daytime earnings and fluctuations; third, without taking into account the length of the overnight period, Different types of overnight information have different predictive ability, it is necessary to classify overnight information. Fourthly, there is leverage effect in stock market, and the leverage effect of intraday information and overnight information is different. For daytime information, negative shocks cause greater volatility, and both positive and negative shocks can increase volatility. For nighttime information, the market is different, and the types of overnight information are different, so the impact of capturing overnight information on intraday fluctuations is not the same.
【学位授予单位】：东北大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F832.51

【相似文献】